Chapter 3
votes and money. Finally, they assume that the higher the barriers to
exit, the greater the incentive for the industry to invest resources in protection. T h e authors suggest that these variables explain the pattern
of protection over time: institutionalised protection for large
industries with high barriers to exit, cyclical protection for
industries with high barriers to entry, and declining protection for
small industries with low entry and exit barriers.
Applicability to the Australian situation
T h e types of models described above do not include public policy as a systematic structure affecting incentives and disincentives facing
firms. This problem is acutely demonstrated in the Australian situation, as Gregory's critique of Anderson's work shows. Anderson's approach was
criticised by Gregory on the grounds that variations in effective
assistance rates across manufacturing industries classified according to 4-digit A S I C categories did not 'focus on the essential nature of tariff setting and tariff changes in this country' (Gregory 1980: 36).
Gregory pointed out that the objectives and operating procedures of the
Tariff Board constituted a 'made to measure' tariff setting procedure (Corden had also made this point). Therefore, tariff levels would
describe the compsirative advantage and disadvantage of the Australian manufacturing sector and would be higher for industries with low wages,
a higher share of w a g e s in value a d d e d , low capital intensity and low
exports.
Glezer's work on interest group formation confirms the shaping role of tariff policy (Glezer 1982: 221). Glezer sdso notes the industry
fragmentation which protection policy induced; the relatively greater
cohesion among smsdl product and industry groups; the influence of 'dominants' in capital-intensive industries such as chemicals and paper
a n d the differing reactions of firms to federal governments' changed perceptions of their usefulness to the economy (Glezer 1982: 222-224;
232-237).
While the role of policy in determining interest has been demonstrated empirically, it has proved difficult to generalise from these
observations. Lowi's work remains the most concerted attempt to generate testable hypotheses (Lowi 1964; 1970). However, Lowi's hypotheses are not testable in the form in which he proposed them because of
difficulties in assigning a n y given policy to the 'regulatory',
'distributive' or 'redistributive' categories (Greenberg et al 1977: 1534). It could be argued that public choice theory has, in fact, been able to encapsulate Lowi's separate categories u n d e r the one heading of regulation.
Nevertheless, within one policy domain (as in the awarding of assistance to manufacturing firms), it is clear that government-determined
incentives and disincentives must be added to economic structural variables, electored factors and general economic conditions in
determining which industries get what sort of assistance and w h e n . The objective of this section is to determine, by an inductive analysis of
the kinds of decisions affecting particular industry sectors, the various factors which contribute to political effectiveness.
Determineuats of political effectiveness
What constitutes political effectiveness, and how can it be detected?
For an industry, achieving favourable regulation by government is an obvious indication of political effectiveness. But not all regulation is
measurable statistically. Moreover, the political pressure required to avoid unfavourable regulation may not be symmetrical with that required
to obtain favourable regulation.
For an industry to be politically effective, its problems must come
within the scope of government action. If an industry believes imports are a problem, it is generally administratively feasible to tax imports.
Chapter 3
But if poor seasonal conditions and depressed export markets are the problem (as was the case in the meat processing industry in the early 1980s) there is little action which can be taken to maintain economic activity.
A higher average rate of industry assistance is not, per se, an
indication that one industry is more politically effective than another. Within most industry groups, the apparent dispersion of effective rates increases as the level of aggregation decreases. As the lAC noted in its
1986 report into the chemicals and plastics industries, there were significant differences in levels of assistance even within 4-digit ASIC levels (lAC 1986a: 38).
In a predominantly needs-based tariff-setting system, there will be a number of factors determining the level of tariff that is set for
particular products: whether the product is locally made or not; whether the product is itself an assembly of smaller components; the degree of packaging it requires; its value; technical characteristics; auid the cost competitiveness of imports relative to domestic production. We
cannot, for example, conclude from the fact that in 1976-77, paper bags had an effective rate of 44% and pulp and paper board an effective rate of 10%, that producers of the first product were more politically
effective than producers of the second.
However, tariff rates worked out at different ASIC levels give only part of the picture. The basic, conscious unit of economic activity in any industry is the firm, and firms are variously arranged around ASIC levels. As well, protection is granted, not by ASIC category, but according to product divisions of the customs tariff.
The political concept of an industry is not a straightforward one. For example, the textiles, clothing and footwear industries are considered, administratively, as one group, and the extent of the employment they represent - that is, their poUtical visibility - is based on this